HopxSAFE
Execute code in isolated cloud containers. Supports Python, JavaScript, Bash, and Go.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Model Context Protocol (MCP) server for HOPX. Enables AI assistants to execute code in isolated cloud containers.
MCP Name: io.github.hopx-ai/hopx-mcp
[](LICENSE) [](https://python.org) [](https://modelcontextprotocol.io)
Installation
uvx hopx-mcp
Get your API key at hopx.ai.
Capabilities
- Execute Python, JavaScript, Bash, and Go in isolated containers
- Data analysis with pandas, numpy, matplotlib (pre-installed)
- File operations and system commands
- Background processes and long-running tasks
Containers auto-destroy after use.
Configuration
Get API Key
Sign up at hopx.ai.
Configure IDE
Add MCP server configuration to your IDE:
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"hopx-sandbox": {
"command": "uvx",
"args": ["hopx-mcp"],
"env": {
"HOPX_API_KEY": "your-api-key-here"
}
}
}
}Restart Claude Desktop after configuration.
Cursor
Add to .cursor/mcp.json:
{
"mcpServers": {
"hopx-sandbox": {
"command": "uvx",
"args": ["hopx-mcp"],
"env": {
"HOPX_API_KEY": "your-api-key-here"
}
}
}
}VS Code
Add to .vscode/mcp.json:
{
"mcpServers": {
"hopx-sandbox": {
"command": "uvx",
"args": ["hopx-mcp"],
"env": {
"HOPX_API_KEY": "your-api-key-here"
}
}
}
}Other IDEs
Configuration format
b692fdef4e22OBSERVED · 2026-10-07Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add hopx-mcp --env API_KEY=${API_KEY} -- uvx hopx-mcp{
"mcpServers": {
"hopx-mcp": {
"command": "uvx",
"args": [
"hopx-mcp"
],
"env": {
"API_KEY": "${API_KEY}"
}
}
}
}Exposed tools (35)
18 read · 12 write · 5 destructive. Blast radius: 5 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
cache_clear | destructive | |
cache_stats | read | |
create_sandbox | write | |
delete_sandbox | destructive | |
env_clear | destructive | |
env_get | read | |
env_set | write | |
execute_code | write | |
execute_code_background | write | |
execute_code_isolated | write | |
execute_code_persistent | write | |
execute_code_rich | write | |
execute_kill_process | destructive | |
execute_list_processes | write | |
file_exists | read | |
file_list | read | |
file_mkdir | read | |
file_read | read | |
file_remove | destructive | |
file_write | write | |
get_agent_url | read | |
get_preview_url | read | |
get_sandbox | read | |
get_system_metrics | read | |
get_template | read | |
get_vm_info | read | |
health | read | GET /health — public health check. |
list_processes | read | |
list_sandboxes | read | |
list_templates | read | |
ping_vm | read | |
resume_sandbox | read | |
run_command | write | |
run_command_background | write | |
update_sandbox_timeout | write |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (1 observation(s))
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (3)
cache_clear, delete_sandbox, env_clear, execute_kill_process, file_remove
.env.sample
Gates applied: no_behavioural_pass, no_license.
b692fdef4e22full audit observations/trust-audit/mcp-server/hopx-ai__hopx.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-07 | b692fdef4e22 | SAFE | B | 89 | first audit |
Questions
What is the Hopx MCP server?
Execute code in isolated cloud containers. Supports Python, JavaScript, Bash, and Go.
What tools does Hopx expose?
35 in total: 18 read-only, 12 that write, and 5 that can delete or overwrite (cache_clear, delete_sandbox, env_clear, execute_kill_process, file_remove). Every one is listed on this page with its risk.
Is Hopx safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 5 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does Hopx need?
It reads API_KEY from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How current is this page?
The grade is for one exact copy of the source (b692fdef4e22), read on 2026-10-07. The repository is watched and re-audited when it changes.